def populate_go_annotations(totalAnnotations,session,engine):
"""
read the annotation file into a dictionary
This will take some time
This function is intended for use with
http://www.geneontology.org/GO.format.gaf-2_0.shtml
"""
timeStart = time.time()
config = Configure()
taxaList = config.log['taxa']
toAdd = []
annotationFile = get_annotation_file()
annotationFid = open(annotationFile,'rU')
wayPoints = [round(int(w)) for w in np.linspace(0,totalAnnotations,20)]
annotationCount = 0
print("...loading mappers")
termIdMap = goterm_mapper(session)
taxaIdMap = taxa_mapper(session)
uniprotIdMap = uniprot_mapper(session)
print("...populating rows")
def queue_entry(goId,evidenceCode,pubmedRefs,uniprotId,geneId,taxon,toAdd,mapper,ignoredAnnotations):
## remove invalid term ids
if not termIdMap.has_key(goId):
queryTerm = session.query(GoTerm).filter_by(alternate_id=goId).first()
if queryTerm == None:
return
go_db_id = queryTerm.id
else:
go_db_id = termIdMap[goId]
## remove invalid uniprot ids
if uniprotId and not mapper.has_key(uniprotId):
return
if uniprotId:
uniprot_db_id = mapper[uniprotId]
else:
uniprot_db_id = None
## remove invalid gene ids
if geneId and not mapper.has_key(geneId):
return
if geneId:
gene_db_id = mapper[geneId]
else:
gene_db_id = None
## ignore annotations that have an outdated taxon
if not taxaIdMap.has_key(taxon):
ignoredAnnotations += 1
return
## get the taxa foreign key
taxon_db_id = taxaIdMap[taxon]
toAdd.append({'go_term_id':go_db_id,'evidence_code':evidenceCode,
'pubmed_refs':pubmedRefs,'uniprot_id':uniprot_db_id,
'gene_id':gene_db_id,'taxa_id':taxon_db_id})
## add annotations from uniprot annotation file
ignoredAnnotationsUniprot = 0
print("...getting annotations from gene_association (uniprot)")
for record in annotationFid:
record = record[:-1].split("\t")
## check that it is a uniprot entry
if record[0][0] == "!":
continue
if record[0] != 'UniProtKB':
continue
uniprotId = record[1]
dbObjectSymbol = record[2]
goId = record[4]
pubmedRefs = record[5]
evidenceCode = record[6]
aspect = record[8]
uniprotEntry = record[10]
goTermName = record[11]
taxon = re.sub("taxon:","",record[12])
date = record[13]
assignedBy = record[14]
if taxon not in taxaList:
continue
## parse the uniprot Entry
if re.search("\|",uniprotEntry):
uniprotEntry = re.split("\|",uniprotEntry)[0]
## ignore annotations with multiple species
if re.search("\|",taxon):
continue
## update progress
annotationCount += 1
if annotationCount in wayPoints:
print("\t%s / %s"%(annotationCount,totalAnnotations))
queue_entry(goId,evidenceCode,pubmedRefs,uniprotEntry,None,taxon,toAdd,
uniprotIdMap,ignoredAnnotationsUniprot)
if len(toAdd) >= 100000: # 100000
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().
values(toAdd))
toAdd = []
print('committing final changes...')
print('ignored annotations after uniprot... %s'%(ignoredAnnotationsUniprot))
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().
values(toAdd))
del uniprotIdMap
annotationFid.close()
## add annotations from gene2go
gene2goFile = get_gene2go_file()
gene2goFid = open(gene2goFile,'rU')
ignoredAnnotationsGene = 0
print("...getting annotations from gene2go")
header = gene2goFid.next()
geneIdMap = gene_mapper(session)
toAdd = []
for record in gene2goFid:
record = record.rstrip("\n")
record = record.split("\t")
if re.search("^\#",record[0]) or len(record) != 8:
continue
taxon = record[0]
ncbiId = record[1]
goId = record[2]
evidenceCode = record[3]
qualifier = record[4]
go_term_description = record[5]
pubmedRefs = record[6]
go_aspect = record[7]
annotationCount += 1
if taxon not in taxaList:
continue
if annotationCount in wayPoints:
print("\t%s / %s"%(annotationCount,totalAnnotations))
queue_entry(goId,evidenceCode,pubmedRefs,None,ncbiId,taxon,toAdd,
geneIdMap,ignoredAnnotationsGene)
if len(toAdd) >= 100000: #100000
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().
values(toAdd))
toAdd = []
print('ignored annotations after gene2go... %s'%(ignoredAnnotationsGene))
print('committing final changes...')
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().
values(toAdd))
timeStr = "...total time taken: %s"%time.strftime('%H:%M:%S', time.gmtime(time.time()-timeStart))
addedStr = "...%s unique go annotation entries were added."%annotationCount
return timeStr,addedStr,(ignoredAnnotationsUniprot,ignoredAnnotationsGene)
def populate_go_annotations(totalAnnotations, session, engine):
"""
read the annotation file into a dictionary
This will take some time
This function is intended for use with
http://www.geneontology.org/GO.format.gaf-2_0.shtml
"""
timeStart = time.time()
config = Configure()
taxaList = config.log['taxa']
toAdd = []
annotationFile = get_annotation_file()
annotationFid = open(annotationFile, 'rU')
wayPoints = [round(int(w)) for w in np.linspace(0, totalAnnotations, 20)]
annotationCount = 0
print("...loading mappers")
termIdMap = goterm_mapper(session)
taxaIdMap = taxa_mapper(session)
uniprotIdMap = uniprot_mapper(session)
print("...populating rows")
def queue_entry(goId, evidenceCode, pubmedRefs, uniprotId, geneId, taxon,
toAdd, mapper, ignoredAnnotations):
## remove invalid term ids
if not termIdMap.has_key(goId):
queryTerm = session.query(GoTerm).filter_by(
alternate_id=goId).first()
if queryTerm == None:
return
go_db_id = queryTerm.id
else:
go_db_id = termIdMap[goId]
## remove invalid uniprot ids
if uniprotId and not mapper.has_key(uniprotId):
return
if uniprotId:
uniprot_db_id = mapper[uniprotId]
else:
uniprot_db_id = None
## remove invalid gene ids
if geneId and not mapper.has_key(geneId):
return
if geneId:
gene_db_id = mapper[geneId]
else:
gene_db_id = None
## ignore annotations that have an outdated taxon
if not taxaIdMap.has_key(taxon):
ignoredAnnotations += 1
return
## get the taxa foreign key
taxon_db_id = taxaIdMap[taxon]
toAdd.append({
'go_term_id': go_db_id,
'evidence_code': evidenceCode,
'pubmed_refs': pubmedRefs,
'uniprot_id': uniprot_db_id,
'gene_id': gene_db_id,
'taxa_id': taxon_db_id
})
## add annotations from uniprot annotation file
ignoredAnnotationsUniprot = 0
print("...getting annotations from gene_association (uniprot)")
for record in annotationFid:
record = record[:-1].split("\t")
## check that it is a uniprot entry
if record[0][0] == "!":
continue
if record[0] != 'UniProtKB':
continue
uniprotId = record[1]
dbObjectSymbol = record[2]
goId = record[4]
pubmedRefs = record[5]
evidenceCode = record[6]
aspect = record[8]
uniprotEntry = record[10]
goTermName = record[11]
taxon = re.sub("taxon:", "", record[12])
date = record[13]
assignedBy = record[14]
if taxon not in taxaList:
continue
## parse the uniprot Entry
if re.search("\|", uniprotEntry):
uniprotEntry = re.split("\|", uniprotEntry)[0]
## ignore annotations with multiple species
if re.search("\|", taxon):
continue
## update progress
annotationCount += 1
if annotationCount in wayPoints:
print("\t%s / %s" % (annotationCount, totalAnnotations))
queue_entry(goId, evidenceCode, pubmedRefs, uniprotEntry, None, taxon,
toAdd, uniprotIdMap, ignoredAnnotationsUniprot)
if len(toAdd) >= 100000: # 100000
with engine.begin() as connection:
connection.execute(
GoAnnotation.__table__.insert().values(toAdd))
toAdd = []
print('committing final changes...')
print('ignored annotations after uniprot... %s' %
(ignoredAnnotationsUniprot))
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().values(toAdd))
del uniprotIdMap
annotationFid.close()
## add annotations from gene2go
gene2goFile = get_gene2go_file()
gene2goFid = open(gene2goFile, 'rU')
ignoredAnnotationsGene = 0
print("...getting annotations from gene2go")
header = gene2goFid.next()
geneIdMap = gene_mapper(session)
toAdd = []
for record in gene2goFid:
record = record.rstrip("\n")
record = record.split("\t")
if re.search("^\#", record[0]) or len(record) != 8:
continue
taxon = record[0]
ncbiId = record[1]
goId = record[2]
evidenceCode = record[3]
qualifier = record[4]
go_term_description = record[5]
pubmedRefs = record[6]
go_aspect = record[7]
annotationCount += 1
if taxon not in taxaList:
continue
if annotationCount in wayPoints:
print("\t%s / %s" % (annotationCount, totalAnnotations))
queue_entry(goId, evidenceCode, pubmedRefs, None, ncbiId, taxon, toAdd,
geneIdMap, ignoredAnnotationsGene)
if len(toAdd) >= 100000: #100000
with engine.begin() as connection:
connection.execute(
GoAnnotation.__table__.insert().values(toAdd))
toAdd = []
print('ignored annotations after gene2go... %s' % (ignoredAnnotationsGene))
print('committing final changes...')
with engine.begin() as connection:
connection.execute(GoAnnotation.__table__.insert().values(toAdd))
timeStr = "...total time taken: %s" % time.strftime(
'%H:%M:%S', time.gmtime(time.time() - timeStart))
addedStr = "...%s unique go annotation entries were added." % annotationCount
return timeStr, addedStr, (ignoredAnnotationsUniprot,
ignoredAnnotationsGene)
def populate_uniprot_table(lineCount,session,engine):
"""
populate the uniprot table with entries from idmappings
"""
config = Configure()
taxaList = config.log['taxa']
timeStart = time.time()
totalLines,totalRecords = 0,0
idmappingFile = get_idmapping_file()
idmappingFid = open(idmappingFile,'rb')
reader = csv.reader(idmappingFid,delimiter="\t")
ac2kbMap,toAdd = {},{}
wayPoints = [round(int(w)) for w in np.linspace(0,lineCount,20)]
print("getting mappers...")
geneIdMap = gene_mapper(session)
taxonIdMap = taxa_mapper(session)
print("mappers loaded... %s"%time.strftime('%H:%M:%S',time.gmtime(time.time()-timeStart)))
def queue_entries(toAdd,geneIdMap,taxonIdMap,engine):
toCommit = []
for uniprotKbEntry, entry in toAdd.iteritems():
db_gene_id = None
db_taxa_id = None
db_gene_taxa_id = None
## convert the gene id to a database key (check old names if we cannot find it)
if entry['gene-id'] == None:
pass
elif geneIdMap.has_key(entry['gene-id']):
db_gene_id = geneIdMap[entry['gene-id']]
elif not geneIdMap.has_key(entry['gene-id']):
_geneIds = [re.sub("\s+","",_ncid) for _ncid in entry['gene-id'].split(";")]
db_gene_id = None
for _gid in _geneIds:
if geneIdMap.has_key(_gid):
db_gene_id= _gid
## convert the taxa id to a database key
if entry['ncbi-taxa-id'] and taxonIdMap.has_key(entry['ncbi-taxa-id']):
db_taxa_id = taxonIdMap[entry['ncbi-taxa-id']]
## check that the linked gene taxa is the same as the entry taxa
#if db_gene_id:
# db_gene_taxa_id = session.query(Gene).filter_by(id=db_gene_id).first().taxa_id
#if db_taxa_id and db_gene_id:
# if db_taxa_id != db_gene_taxa_id:
# print("WARNING: two taxa present in single uniprot entry? %s %s "%(uniprotKbEntry,\
# entry['gene-id']))
## if no taxa was provdied use the one assocated with the linked gene
#if not db_taxa_id:
# dgeneQuery = session.query(Gene).filter_by(id=db_gene_id).first()
# if dgeneQuery:
# db_taxa_id = dgeneQuery.taxa_id
## ensure we are in appropriate taxa
if entry['ncbi-taxa-id'] not in taxaList:
continue
## ready the uniprot-ac and refseq rows
entry['uniprot-ac'] = list(entry['uniprot-ac'])
if len(entry['uniprot-ac']) == 0:
entry['uniprot-ac'] = None
elif len(entry['uniprot-ac']) == 1:
entry['uniprot-ac'] = entry['uniprot-ac'][0]
else:
entry['uniprot-ac'] = ";".join(entry['uniprot-ac'])
entry['refseq'] = list(entry['refseq'])
if len(entry['refseq']) == 0:
entry['refseq'] = None
elif len(entry['refseq']) == 1:
entry['refseq'] = entry['refseq'][0]
else:
entry['refseq'] = ";".join(entry['refseq'])
## commit to db
toCommit.append({'uniprot_ac':entry['uniprot-ac'],'uniprot_entry':uniprotKbEntry,
'refseq':entry['refseq'],'taxa_id':db_taxa_id,'gene_id':db_gene_id})
if len(toCommit) > 0:
with engine.begin() as connection:
connection.execute(Uniprot.__table__.insert().
values(toCommit))
## parse the idmapping file into the db
for record in reader:
if len(record) != 3:
continue
uniprotKbAc,uniprotKbEntry,ncbiId,refseq,ncbiTaxaId = None,None,None,None,None
uniprotKbAc = record[0]
totalLines += 1
if totalLines in wayPoints:
print("\t%s / %s"%(totalLines,lineCount))
if record[1] == 'NCBI_TaxID':
ncbiTaxaId = record[2]
elif record[1] == 'GeneID':
ncbiId = record[2]
elif record[1] == 'UniProtKB-ID':
uniprotKbEntry = record[2]
if not ac2kbMap.has_key(uniprotKbAc):
ac2kbMap[uniprotKbAc] = uniprotKbEntry
elif record[1] == 'RefSeq':
refseq = record[2]
else:
continue
## skip the XXXX-1 like uniprot ac
if ac2kbMap.has_key(uniprotKbAc) == False:
continue
## get current key
uniprotKbEntry = ac2kbMap[uniprotKbAc]
## make new entry if necessary
if uniprotKbEntry and not toAdd.has_key(uniprotKbEntry):
## queue entries in blocks
totalRecords += 1
if totalRecords % 100000 == 0:
queue_entries(toAdd,geneIdMap,taxonIdMap,engine)
toAdd,ac2kbMap = {},{}
ac2kbMap[uniprotKbAc] = uniprotKbEntry
toAdd[uniprotKbEntry] = {'ncbi-taxa-id':None,
'gene-id':None,
'uniprot-ac':set([]),
'refseq':set([])}
## populate uniprot dictionary
toAdd[uniprotKbEntry]['uniprot-ac'].update([uniprotKbAc])
if ncbiTaxaId:
toAdd[uniprotKbEntry]['ncbi-taxa-id'] = ncbiTaxaId
elif ncbiId:
toAdd[uniprotKbEntry]['gene-id'] = ncbiId
elif refseq:
toAdd[uniprotKbEntry]['refseq'].update([refseq])
## queue any remaining
if len(toAdd.keys()) > 0:
queue_entries(toAdd,geneIdMap,taxonIdMap,engine)
## clean up
idmappingFid.close()
timeStr = "...total time taken: %s"%time.strftime('%H:%M:%S', time.gmtime(time.time()-timeStart))
addedStr = "...%s unique uniprot entries were added."%totalRecords
return timeStr,addedStr
def populate_uniprot_table(lineCount, session, engine):
"""
populate the uniprot table with entries from idmappings
"""
config = Configure()
taxaList = config.log['taxa']
timeStart = time.time()
totalLines, totalRecords = 0, 0
idmappingFile = get_idmapping_file()
idmappingFid = open(idmappingFile, 'rb')
reader = csv.reader(idmappingFid, delimiter="\t")
ac2kbMap, toAdd = {}, {}
wayPoints = [round(int(w)) for w in np.linspace(0, lineCount, 20)]
print("getting mappers...")
geneIdMap = gene_mapper(session)
taxonIdMap = taxa_mapper(session)
print("mappers loaded... %s" %
time.strftime('%H:%M:%S', time.gmtime(time.time() - timeStart)))
def queue_entries(toAdd, geneIdMap, taxonIdMap, engine):
toCommit = []
for uniprotKbEntry, entry in toAdd.iteritems():
db_gene_id = None
db_taxa_id = None
db_gene_taxa_id = None
## convert the gene id to a database key (check old names if we cannot find it)
if entry['gene-id'] == None:
pass
elif geneIdMap.has_key(entry['gene-id']):
db_gene_id = geneIdMap[entry['gene-id']]
elif not geneIdMap.has_key(entry['gene-id']):
_geneIds = [
re.sub("\s+", "", _ncid)
for _ncid in entry['gene-id'].split(";")
]
db_gene_id = None
for _gid in _geneIds:
if geneIdMap.has_key(_gid):
db_gene_id = _gid
## convert the taxa id to a database key
if entry['ncbi-taxa-id'] and taxonIdMap.has_key(
entry['ncbi-taxa-id']):
db_taxa_id = taxonIdMap[entry['ncbi-taxa-id']]
## check that the linked gene taxa is the same as the entry taxa
#if db_gene_id:
# db_gene_taxa_id = session.query(Gene).filter_by(id=db_gene_id).first().taxa_id
#if db_taxa_id and db_gene_id:
# if db_taxa_id != db_gene_taxa_id:
# print("WARNING: two taxa present in single uniprot entry? %s %s "%(uniprotKbEntry,\
# entry['gene-id']))
## if no taxa was provdied use the one assocated with the linked gene
#if not db_taxa_id:
# dgeneQuery = session.query(Gene).filter_by(id=db_gene_id).first()
# if dgeneQuery:
# db_taxa_id = dgeneQuery.taxa_id
## ensure we are in appropriate taxa
if entry['ncbi-taxa-id'] not in taxaList:
continue
## ready the uniprot-ac and refseq rows
entry['uniprot-ac'] = list(entry['uniprot-ac'])
if len(entry['uniprot-ac']) == 0:
entry['uniprot-ac'] = None
elif len(entry['uniprot-ac']) == 1:
entry['uniprot-ac'] = entry['uniprot-ac'][0]
else:
entry['uniprot-ac'] = ";".join(entry['uniprot-ac'])
entry['refseq'] = list(entry['refseq'])
if len(entry['refseq']) == 0:
entry['refseq'] = None
elif len(entry['refseq']) == 1:
entry['refseq'] = entry['refseq'][0]
else:
entry['refseq'] = ";".join(entry['refseq'])
## commit to db
toCommit.append({
'uniprot_ac': entry['uniprot-ac'],
'uniprot_entry': uniprotKbEntry,
'refseq': entry['refseq'],
'taxa_id': db_taxa_id,
'gene_id': db_gene_id
})
if len(toCommit) > 0:
with engine.begin() as connection:
connection.execute(Uniprot.__table__.insert().values(toCommit))
## parse the idmapping file into the db
for record in reader:
if len(record) != 3:
continue
uniprotKbAc, uniprotKbEntry, ncbiId, refseq, ncbiTaxaId = None, None, None, None, None
uniprotKbAc = record[0]
totalLines += 1
if totalLines in wayPoints:
print("\t%s / %s" % (totalLines, lineCount))
if record[1] == 'NCBI_TaxID':
ncbiTaxaId = record[2]
elif record[1] == 'GeneID':
ncbiId = record[2]
elif record[1] == 'UniProtKB-ID':
uniprotKbEntry = record[2]
if not ac2kbMap.has_key(uniprotKbAc):
ac2kbMap[uniprotKbAc] = uniprotKbEntry
elif record[1] == 'RefSeq':
refseq = record[2]
else:
continue
## skip the XXXX-1 like uniprot ac
if ac2kbMap.has_key(uniprotKbAc) == False:
continue
## get current key
uniprotKbEntry = ac2kbMap[uniprotKbAc]
## make new entry if necessary
if uniprotKbEntry and not toAdd.has_key(uniprotKbEntry):
## queue entries in blocks
totalRecords += 1
if totalRecords % 100000 == 0:
queue_entries(toAdd, geneIdMap, taxonIdMap, engine)
toAdd, ac2kbMap = {}, {}
ac2kbMap[uniprotKbAc] = uniprotKbEntry
toAdd[uniprotKbEntry] = {
'ncbi-taxa-id': None,
'gene-id': None,
'uniprot-ac': set([]),
'refseq': set([])
}
## populate uniprot dictionary
toAdd[uniprotKbEntry]['uniprot-ac'].update([uniprotKbAc])
if ncbiTaxaId:
toAdd[uniprotKbEntry]['ncbi-taxa-id'] = ncbiTaxaId
elif ncbiId:
toAdd[uniprotKbEntry]['gene-id'] = ncbiId
elif refseq:
toAdd[uniprotKbEntry]['refseq'].update([refseq])
## queue any remaining
if len(toAdd.keys()) > 0:
queue_entries(toAdd, geneIdMap, taxonIdMap, engine)
## clean up
idmappingFid.close()
timeStr = "...total time taken: %s" % time.strftime(
'%H:%M:%S', time.gmtime(time.time() - timeStart))
addedStr = "...%s unique uniprot entries were added." % totalRecords
return timeStr, addedStr